在 X 上的一篇帖子中：https://x.com/merettm/status/2095023204993490967，OpenAI 首席科学家 Jakub Pachocki 强调了实验室对可读思维链的承诺。Pachocki 写道：“自从我们最早的推理模型开始，OpenAI 就一直致力于保存和利用思维链监控。这是我们当前研究计划的核心目标。”
所有 AI 模型都进行一定程度的不透明推理，很少有研究人员将思维链日志视为模型推理的直接表现。然而，这些提示并不能消除人们对不透明递归可能使 AI 推理更难监控的担忧，尤其当它在不同模型中越来越广泛使用时。在周三早上的后续报道中，《The Information》报道称，Anthropic 和 Google DeepMind 已经在讨论该技术。
在回应该新闻的一篇帖子中：https://x.com/ryangreenblatt/status/2094996656186081642?s=46&t=45_xAnRsdQP1GVqYv9Gdbw，Redwood Research 首席科学家 Ryan Greenblatt 表示，不透明推理可能比传统链式思维推理增长得更快，有效地将所有推理从可见渠道中移除。
微软测试解决最新长时间Outlook宕机的方法：https://techcrunch.com/2026/08/31/microsoft-tests-fix-for-latest-hours-long-outlook-outage/ Sarah Perez：https://techcrunch.com/author/sarah-perez/
MapQuest应用在拒绝重命名安大略湖后跃升至导航类第一：https://techcrunch.com/2026/08/31/mapquests-app-surges-to-no-1-in-navigation-after-refusing-to-rename-lake-ontario/ Sarah Perez：https://techcrunch.com/author/sarah-perez/
OpenAI’s new Astra model will use a reasoning technique called “recurrent depth” that allows it to operate outside of the sequential thinking that characterizes most reasoning models, the The Information reported：https://www.theinformation.com/articles/secret-technique-behind-openais-astra-model-sparks-security-concerns on Tuesday. This technique, also called “opaque recurrence,” will likely make the model’s chain of thought more difficult to monitor — and that has AI safety experts rattled.
While Astra’s use of the technique is reportedly limited, its emergence has still raised significant concerns among AI safety experts.
“I am extremely concerned by the reporting that Astra uses opaque recurrence,” wrote Redwood CEO Buck Shlegeris in a post：https://x.com/bshlgrs/status/2094990313513439464?s=46&t=45_xAnRsdQP1GVqYv9Gdbw after the news broke. “I don’t know whether Astra is much less CoT monitorable than previous models. But if OpenAI pushes this technique further, they’ll have the option to massively increase the recurrence and totally destroys CoT monitorability.”
Longtime AI safety advocate Zvi Mowshowitz also weighed in and wrote：https://thezvi.substack.com/p/anthropic-has-some-alignment-problems?open=false#%C2%A7this-just-in that laws might be necessary to prevent a “race to the bottom” among AI labs.
“The technique is playing with fire, risking a taboo that OpenAI and Anthropic have fought to establish that we work hard to maintain Chain of Thought faithfulness and monitorability for as long as we can,” Mowshowitz wrote. “More intensive use of such techniques would probably damage monitorability.”
Under normal circumstances, a reasoning model’s chain of thought provides the sequential steps taken by the model as it attempts to solve a problem. While the representation is imperfect, it still serves as a valuable tool for monitoring misbehavior or misalignment. In the case of OpenAI’s recent rogue agent activity, chain-of-thought records were an important tool in teasing out why agents behaved the way they did.
In opaque recurrence, the model takes a less linear approach, processing the same query several times in a loop. The result leaves fewer legible traces, effectively side-stepping a conventional chain-of-thought record.
Crucially, Astra’s use of the technique appears to be limited. The model’s chain of thought is still expected to be legible, and the company pushed back against any suggestion that it would shift to “neuralese.” OpenAI has already announced plans for extensive chain-of-thought monitoring systems as part of its forward-looking safety plans.
In a post on X：https://x.com/merettm/status/2095023204993490967, OpenAI chief scientist Jakub Pachocki emphasized the lab’s commitment to legible chains of thought. “OpenAI has worked to preserve and utilize chain-of-thought monitoring since our very first reasoning models,” Pachocki wrote. “It’s a core goal of our current research program.
All AI models do some quantity of opaque reasoning, and few researchers take chain-of-thought logs as a direct representation of a model’s reasoning. Still, those caveats don’t dispel the concern that opaque recurrence may make AI reasoning harder to monitor, particularly as it grows in use across different models. In a follow-up report Wednesday morning, The Information reported that both Anthropic and Google DeepMind were already discussing the technique.
In a post responding to the news,：https://x.com/ryangreenblatt/status/2094996656186081642?s=46&t=45_xAnRsdQP1GVqYv9Gdbw Redwood Research chief scientist Ryan Greenblatt said opaque reasoning could easily scale faster than conventional chain-of-thought reasoning, effectively removing all reasoning from visible channels.
“My biggest concern is that a natural progression from here would involve scaling up the opaque reasoning to the point where the model reasons entirely or almost entirely in latent space,” Greenblatt wrote. “I hope it isn’t too late to avoid the most concerning architectures and that OpenAI will stop here.”
When you purchase through links in our articles, we may earn a small commission：https://techcrunch.com/techcrunch-affiliate-monetization-standards/. This doesn’t affect our editorial independence.
Don’t miss out . The startup community will gather to answer a pivotal question: How do you build sustainably in the AI era?
Uber is laying off 10% of staff, or 3,300 people：https://techcrunch.com/2026/09/02/uber-is-laying-off-10-of-staff-or-3300-people/ Ram Iyer：https://techcrunch.com/author/ram-iyer/
Apple shares ‘shocking evidence’ against former employee accused of stealing company data for OpenAI：https://techcrunch.com/2026/08/31/apple-shares-shocking-evidence-against-former-employee-accused-of-stealing-company-data-for-openai/ Amanda Silberling：https://techcrunch.com/author/amanda-silberling/
Microsoft tests fix for latest hours-long Outlook outage：https://techcrunch.com/2026/08/31/microsoft-tests-fix-for-latest-hours-long-outlook-outage/ Sarah Perez：https://techcrunch.com/author/sarah-perez/
MapQuest’s app surges to No. 1 in Navigation after refusing to rename Lake Ontario：https://techcrunch.com/2026/08/31/mapquests-app-surges-to-no-1-in-navigation-after-refusing-to-rename-lake-ontario/ Sarah Perez：https://techcrunch.com/author/sarah-perez/
Musk’s faster path to more gas turbines comes with pollution problem：https://techcrunch.com/2026/08/30/musks-faster-path-to-more-gas-turbines-comes-with-pollution-problem/ Connie Loizos：https://techcrunch.com/author/connie-loizos/
Hugging Face is selling a cute $399 open source duck robot, Microduck：https://techcrunch.com/2026/08/27/hugging-face-is-selling-a-cute-399-open-source-duck-robot-microduck/ Rebecca Bellan：https://techcrunch.com/author/rebecca-bellan/
Nvidia closes in on Hugging Face acquisition：https://techcrunch.com/2026/08/26/nvidia-closes-in-on-hugging-face-acquisition/ Connie Loizos：https://techcrunch.com/author/connie-loizos/
扫码直达文章详情